<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
<title></title>
</head>
<body bgcolor="#ffffff" text="#000000">
Unless you would start a compliling split on AMD/Pentium, because you
could make to different compiles for both Athlon and P4, however I
think bugs might be lots harder to track that way.<br>
<br>
<div class="moz-signature"><big><font face="Verdana" size="-2"><big>Mortis<br>
<a href="http://etbf.kicks-ass.net">European Tactical Battlefield Forces</a></big></font></big></div>
<br>
<br>
Kingsley Foreman schreef:
<blockquote cite="mid061701c5aa5c$232a1200$0100a8c0@UGLYPUNK"
type="cite">the only issue with that is the people running athlons, it
could cause issues.
<br>
<br>
kingsley
<br>
<br>
----- Original Message ----- From: "Andreas Andersson"
<a class="moz-txt-link-rfc2396E" href="mailto:andreas.andersson@dice.se"><andreas.andersson@dice.se></a>
<br>
To: <a class="moz-txt-link-rfc2396E" href="mailto:bf1942@icculus.org"><bf1942@icculus.org></a>
<br>
Sent: Friday, August 26, 2005 10:48 PM
<br>
Subject: RE: [bf1942] Are you using the 64-bit server much?
<br>
<br>
<br>
<blockquote type="cite">I should probably up to pentuim4 for the
32-bit binary as well. Pentium3
<br>
doesn't have enough power to handle the bf2-server anyway.
<br>
<br>
/Andreas
<br>
<br>
-----Original Message-----
<br>
From: Andreas Andersson [<a class="moz-txt-link-freetext" href="mailto:andreas.andersson@dice.se">mailto:andreas.andersson@dice.se</a>]
<br>
Sent: den 26 augusti 2005 15:13
<br>
To: <a class="moz-txt-link-abbreviated" href="mailto:bf1942@icculus.org">bf1942@icculus.org</a>
<br>
Subject: RE: [bf1942] Are you using the 64-bit server much?
<br>
<br>
Yeah you might be right...
<br>
<br>
This is what we use right now:
<br>
32bit linux:
<br>
gcc 3.3.5 with -O2 and -march=pentuim3
<br>
<br>
64Bit linux:
<br>
gcc 3.4.3 with -O2 and -m64
<br>
<br>
Windows:
<br>
VC.Net 7.1 with Full optimizations( /Ox)
<br>
<br>
I'm thinking of increasing -O2 to -O3. But I seem to remember that this
<br>
can break introduce bugs? Have you had any experience with this?
<br>
<br>
And maybe it would be a good idea to add a -march setting to the 64-bit
<br>
binary as well. To athlon64 or something.
<br>
<br>
/AndreasA
<br>
<br>
<br>
-----Original Message-----
<br>
From: ScratchMonkey [<a class="moz-txt-link-freetext" href="mailto:ScratchMonkey@MatureAsskickers.net">mailto:ScratchMonkey@MatureAsskickers.net</a>]
<br>
Sent: den 25 augusti 2005 20:09
<br>
To: <a class="moz-txt-link-abbreviated" href="mailto:bf1942@icculus.org">bf1942@icculus.org</a>
<br>
Subject: Re: [bf1942] Are you using the 64-bit server much?
<br>
<br>
--On Thursday, August 25, 2005 1:36 PM +0100 Steven Hartland
<br>
<a class="moz-txt-link-rfc2396E" href="mailto:killing@multiplay.co.uk"><killing@multiplay.co.uk></a> wrote:
<br>
<br>
<blockquote type="cite">The figures we saw where:
<br>
1. a 32bit linux server requires 10% more CPU then a windows one
<br>
2. a 64bit linux server requires 5% more CPU than a 32bit linux one.
<br>
<br>
This really pains me to say but with those figures we have had no
<br>
option but to put 95% of our servers on windows :(
<br>
</blockquote>
<br>
It would be useful to know which compilers, versions, and compiler
<br>
switches
<br>
were used for the 3 versions. I suspect that will have a lot more
<br>
influence
<br>
than kernel version.
<br>
<br>
<br>
<br>
<br>
</blockquote>
<br>
<br>
<br>
</blockquote>
</body>
</html>